User Client

The UserClient is a user-scoped client returned by client.for_user(). All memory operations are automatically scoped to this user.

Creating a user client

user = client.for_user("user-123")

Methods

remember(content)

Store content as memory. Automatically extracts entities and relationships.

# Simple text
result = user.remember("Alice is a software engineer who uses Python")
 
# Multiple items
result = user.remember(["First fact", "Second fact"])
 
# Document object
from memorer import Document
result = user.remember(Document(content="...", metadata={"source": "chat"}))

Returns: IngestResponse with fields:

• entities_created — Number of entities extracted
• relationships_created — Number of relationships created
• episodes_created — Number of temporal episodes created
• processing_time_ms — Processing time in milliseconds
• status — "success"

recall(query, **kwargs)

Query memory using semantic similarity and optional graph reasoning.

results = user.recall("What does Alice do?")
print(results.context)  # Assembled context string for LLM prompts
 
for r in results:
    print(f"{r.content} (score: {r.relevance_score})")

Returns: QueryResponse — iterable of QueryResult objects with a .context property containing assembled text for LLM prompts.

forget(memory_id)

Soft-delete a specific memory.

user.forget("memory-uuid-here")

conversation(session_id)

Get or create a conversation session.

# Create a new conversation
conv = user.conversation()
 
# Resume an existing conversation
conv = user.conversation("existing-conversation-id")
 
conv.add("user", "Hello!")
result = conv.recall("what did we talk about?")

Returns: ConversationClient

Resource Attributes